04. application of sampling phased array technique for ultrasonic inspection of cfrp components.pdf

7/29/2019 04. Application of Sampling Phased Array Technique for Ultrasonic Inspection of CFRP Components.pdf

1/16

Application of Sampling Phased Array

Technique for Ultrasonic Inspection of CFRPComponents

Andrey BULAVINOV, Fraunhofer-IZFP, Saarbrcken, GermanyRandolf HANKE, Fraunhofer-EZRT, Frth, Germany

Julia HEGEMANN, EADS, Munich, GermanyMichael KRNING, Fraunhofer-IZFP, Saarbrcken, GermanyReinhold OSTER, Eurocopter Deutschland, GmbH, Germany

Sergey PUDOVIKOV, Fraunhofer-IZFP, Saarbrcken, GermanyKrishna Mohan REDDY, Lucid Software, Chennai, IndiaRamanan Sridaran VENKAT, Fraunhofer-IZFP, Saarbrcken, Germany

Abstract. The article presents latest results on ultrasonic inspection of carbon-fiberstructures by the Sampling Phased Array method. A comparative study of inspectionresults obtained by common single element transducer and classical Phased Arraymethods is done. The advantages and potential of Sampling Phased Array techniquefor practical use are also discussed.

Introduction

Quality of complex CFRP (CFK) structures is subjected to very high requirements. Onlythe material defects in strictly defined tolerable ranges are allowed in the completed CFK-units. Thus NDT methods have to be applied to assure the quality of components withingiven preferred tolerances. The most applied NDT technique is ultrasonic testing which isan economic and effective testing modality for characterizing material defects.

A cost effective inspection of complex CFK-structures is a challenging task. Incontrast to material testing in other branches (e.g. steel production) aerospace hasinnovative techniques and solutions due to rapid development of manufacturingtechnologies. Thus, the advanced techniques like Phased Arrays are finding increasingapplications in the inspection of aircraft components.

The Sampling Phased Array Technology is the latest advancement in the phasedarray technology, which allows improved defect finding and characterisation of material forfast and the cost-effective quantitative automatic inspection for all applications. The currentarticle presents the latest results of Sampling Phased Array testing of CFK-components.

1. Task definition and solution approaches

1.1 Ultrasonic testing of CFK-components. Characteristic material defects

The ultrasonic testing of CFK-components has a number of distinct challenges whencompared with other applications. Practical use of ultrasonic testing is strongly correlated


2/16

with the defects prevalent for this type of material. The most important of them are asfollows: Delaminations

Delaminations refer to two types - interlaminar (matrix failure) and intralaminar(matrix/fiber failures.

PorosityThe ultrasonic testing of porosity is normally based on measurement of back wall echoamplitude. For the inspection of porosity, a stable back wall echo signal with sufficientsignal to noise ratio has to be ensured.

Plastic or paperfilmsDetection of films can be carried out by measurement of back wall echo attenuation and

by measurement of intermediate echoes in the pulse-echo mode.

Normal probes are often used for ultrasonic testing of CFK-components since thecomponents usually have very small wall thicknesses and due to in-plane orientation ofmaterial defects. For complex shapes, e.g. testing of curved profiles or radii testing, phasedarray transducers can be applied.

1.2 Sampling Phased Array Technique

The Phased Array technology provides test data via an array of individual transducers,which transmit and receive in a phase-controlled fashion. The phase control is via theelectronics and / or software [1]. The implementation of Phased Array systems for materialtesting and evaluation utilizes only a small portion of the overall data acquisition capability,since the acoustic transmissions for specific angle of incidence are time-phased and thereceived signals are then summarized. This means that the entire array acts as a singletransducer in accordance with the Sampling Theorem. However, if the time-domain signalsfrom the individual transducer elements are acquired, the resulting data can then besummarized with arbitrary phase information to permit data processing of all physically

possible angles of incidence and focus points from a single data set. This concept is referredto by Fraunhofer-IZFP as the Sampling Phased Array system [2, 3].

=

N

i

njA1

j=1,2,3,4,5,6..N

Figure 1: Data acquisition and Processing for Sampling Phased Array

Phased ArraySearch Unit

Transmitter i

Transmitting

Receiving

1

2

1

2

i - Transmitterj - Receiver


3/16

The measurement of signals in a time domain Aij(t) with i = j (i: transmitter elementnumber, j: receiver element number) is the essence of the Synthetic Aperture FocusTechnique (SAFT) where one Phased Array element is virtually steered over the aperture ofthe array. The additional acquisition of the time domain signals Aij with i j can be used asa data set for the solution of migration algorithms (e.g. Kirchoff Algorithm), for image

reconstruction, as shown in Figures 1 and described in References [4] and [5]. Thisreconstruction principle is illustrated in Figure 2 below.A significant benefit of this reconstruction technique, referred to by Fraunhofer-

IZFP, as the SynFoc technique is that the small test sampling errors in the algorithmsolution will not adversely affect the image reconstruction.

Figure 2: Image Reconstruction via Kirchoff Algorithm

1.3 Task definition

A set of test specimens with artificial and natural defects was investigated in thisdevelopment project. The objective was to quickly evaluate the suitability of applicable

ultrasonic techniques for detecting characteristic defects in CFK components in anautomatic inspection environment. The immersion technique is preferred for automaticinspection.

The following are selected:1. Traditional Pulse Echo testing with single element transducer.2. Conventional phased array testing.3. Sampling phased array.

Results of these three techniques are compared and discussed in the present article.

Z

x2

z2

x1

z3

z1

x3

Spatial ModelSuperimposition Principle

X

Z

x2

z2

x1

z3

z1

x3

t Data Hyperbola

x2

z2

x1

z3

z1

x3Xx2

z2

x1

z3

Z

z1

x3

Point Reflectors in theReconstruction Volume


4/16

2. Experimental set-up

Ultrasonic measurements were conducted on ultrasonic measuring station that consists ofimmersion tank, 3-axes manipulator (Figure 3a), and 64-channel phased array systemcapable of multichannel inspection as well as Conventional Phased Array and Sampling

Phased Array inspection in contact or immersion technique [6].Double-gimballed probe holders were designed for testing objects with slightly

curved surfaces (Figures 3b and 3c).In Sampling Phased Array mode 3D-images of inspected volume can be

reconstructed in real time. (Figures 3d) [7].a) b)

c)

d)

a) Immersion tank with 3-axes manipulatorb) Double-gimballed probe holder for scanning on slightly curved vertical surfacesc) Double-gimballed probe holder for scanning on slightly curved horizontal surfacesd) Representation of inspection results in 3D-view

Figure 3: Ultrasonic measuring station


5/16

3. Investigation results

3.1 Test specimens

A variety of test specimens were investigated under the scope of the project. Due toconfidentiality requirements, only a few results are presented in this paper.A special step-wedge specimen is used for sensitivity calibration. This type of test

specimen is commonly used for porosity testing. Besides this, the properties of the stepwedge should be equivalent to the material properties of the CFK-structure.Furthermore, the specimen has artificial defects (flat bottom holes) placed close to face andthe back wall for every step. Figure 4 represents the photo of the step wedge.

Figure 4: Step wedge

In the current project, another test specimen with embedded film type defects

(Figure 5) is used for measurement. Unlike delaminations, these defects are embedded intothe structure and are difficult to detect due to poor reflectivity. The wall thickness measuredis 4 mm.

Figure 5: Test specimen with film-like defects


6/16

The third specimen that was tested (Fig.6) represents a CFK-component with aslightly curved surface.

Figure 6: Real CFK component

3.1 Ultrasonic transducers

A single element transducer designed specially for testing of thin CFK-components is usedas a reference for phased array inspection due to its excellent properties for detection andcharacterization of relevant material defects. This probe with the angle of incidence of 0and a wedge of 17 mm is well suited to detect near surface delaminations. This is enabled

by the design of wedge parameters and selection of appropriate crystal size.However, a probe of this type cannot be used for automatic testing since mechanical

contact is not feasible for components with complex geometries or limited accessibility.

Phased Array inspection in immersion technique was favoured as being moresuitable for automatic inspection purposes. The Phased Array probes from Olympus 5L16(pitch 0.6 mm, element length 12 mm) was applied with the water gap of 12 mm.

3.2 Results of ultrasonic testing

Step wedge described in chapter 3.1 is used for sensitivity calibration. TGC curves aremeasured for three modes, single element probe, conventional Phased Array mode andSampling Phased Array mode. The results are shown in Figure 7. In the undisturbed area,all the steps deliver equal back wall echo height and thus represent the same colour in B-

Scan images.


7/16

Figure 7: Sensitivity calibration on step wedge

3.2.1 Detection of delaminations

Step wedge represented in chapter 3.1 has artificial defects FBH 3 and 5 mm placedclose (0.5 mm) to the front and back wall of the specimen.

Measurements with a single element probe from IntelligeNDT, made in contacttechnique provide excellent detection of reflectors up to the depth of 15 mm (Figure 8).With increasing depth the resolution decreases and detectability also becomes reduced.

Figure 8: B-Scan of step wedge with artificial defects measured by single element probe

The results of Conventional Phased Array technique measured at the same area arerepresented in Figure 9. Due to rather low resolution and presence of surface echo thedetectability of near surface delamination can be stated as being very poor.

0

5

10

15

20

25

30

35

40

45

0 5 10 15 20

0

5

10

15

20

25

30

35

40

45

0 5 10 15 20

0

5

10

15

20

25

30

35

40

45

50

0 5 10 15 20

Single

elementprobe

Phased

ArraySamplin

dB dBdBg

Phased

Arraymmmmmm


8/16

Figure 9: B-Scan of step wedge with artificial defects

measured in conventional Phased Array mode

In the Sampling Phased Array mode B-Scan view obtained with measurements from the sametransducer view allows detection of near to surface defects (Figure 10).

Figure 10: B-Scan of step wedge with artificial defects

measured in Sampling Phased Array mode


9/16

4.2.2 Detection of film-like defectsDiscontinuities of this type are usually caused by the so-called human factor whichinclude incorrect operations or operator error.

These defects can be difficult to find since the acoustic impedance mismatch causedby melted film may not be significant. The film type and manufacturing process are criticalfactors that determine detectability. Nevertheless such inclusions can be critical and have to

be detected.The test specimens with different type of embedded films are manufactured to

investigate their detectability by ultrasonic testing (Figure 11). The results achieved byultrasonic testing are rather promising.

Figure 11: Test specimen with film-like defects and its 3D-image obtained by Sampling

Phased Array technique

Figures 12 14 represent inspection results on the specimen by three techniques.Single element probe in contact technique could distinctly find all the artificial defects. Asit can be seen in Figure 12 only monitoring of back wall echo is not sufficient for detectionof film-like defects due to low contrast of indications. The gating out of the back wall echosignal in the C-Scan view increases image contrast and improves detection of film-likedefects.

a) C-Scan with gated in backwall echo b) C-Scan with gated out backwall echo

Figure 12: Detection of film-like defects with single element transducer in contact technique


10/16

Inspection results using conventional Phased Array method in immersion techniqueprovides reasonable detection of film-like defects. However, low sensitivity andresolution and poor signal to noise ratio cannot be overlooked in the C-Scan views(Figure 13).


Figure 13: Detection of film-like defects with PA-transducerin conventional phased array mode in immersion technique

Unlike conventional Phased Array technique the Sampling Phased Array method inimmersion offers high contrast of ultrasonic images (Figure 14). The results are as good asthose with a single element transducer.


Figure 14: Detection of film-like defects with PA-transducer

in Sampling Phased Array mode in immersion technique

4.2.3 Porosity testing on an actual CFK- componentAfter sensitivity calibration for every method (see Figure 6), an actual CFK-component isinspected. The results are represented in Figures 15 17. For each method a C-Scan imageover entire scanning area and a B-Scan image for one index position are represented.


11/16

B Scan plane

a) C-Scan

Changing of wall thickness

B) B-Scan

Figure 15: Ultrasonic testing of an actual CFK component with single element transducer


12/16

B-Scan plane

a) C-Scan

B) B-Scan

Figure 16: Ultrasonic testing of an actual CFK component with conventional phased array technique



13/16

B-Scan plane

Figure 17: Ultrasonic testing of an actual CFK component with Sampling Phased Array technique


a) C-Scan

B) B-Scan


14/16

Comparing these three images we can say that Sampling Phased Array method provideshigher resolution in the steering plane. Even inclined backwall area (see marked area in theB-Scan) can be mapped well, unlike the image by single element probe and conventional

phased array mode.

4. Industrial Implementation

A potential implementation of testing procedure for coplanar and slightly curved CFK-components is shown in Figure 18. A linear Phased Array transducer consisting ofreasonably large number of elements (e.g. up to 128) is moved across its sweeping plane.Several virtual transducers (group of elements functionally associated together for onetransmitter-receiver act) are working in parallel according to Sampling Phased Array

principle. To cover the entire volume in every probe position, several parallel shots may berequired. Thus transmitting and receiving elements have to be multiplexed to build virtual

Sampling Phased Array transducers. The use of parallel transducers ensures that a very highinspection speed can be achieved.

CFK component

Phased Array

transducer

Scanning

trajectory

Figure 18: Principle of fast testing procedure by Sampling Phased Array technique on slightlycurved CFK component

Since an arbitrary angle of incidence can be reconstructed after one shot of virtualSampling Phased Array transducer, a correct backwall echo monitoring for slightly curvedsurface can be implemented for taking into account the local inclination in every probe

position.Nevertheless for exact positioning of ultrasonic transducer with respect to the

component a robotized motion control system is to be integrated for highly reproducibleand reliable inspection. Several synchronized robot arms can simultaneously scan an

inspection object for achieving the shortest required inspection time.

5. Future Directions

5.1Radii inspectionDetection of material discontinuities in areas with strong curvature (radius up to 3 mm) is achallenging task. The best solution for CFK-components with varied or complexgeometries would be testing in immersion technique.

Currently the first simulation (Figure 19) and experiments for practical applicationof Sampling Phased Array method in immersion technique are in progress. The results areexpected in the near future.


15/16

Figure 19: Simulation of radii testing by curved phased array transducer

5.2Automatic 3D evaluation of inspection resultsSampling Phased Array technique provides a fast reconstruction of ultrasonic images in 3-dimensional space (Figure 20). An important task for practical use of this advantage is afully or semi-automated interpretation of ultrasonic images, which provides a maximumsupport to operator by evaluation of inspection results.

This task has to be solved by means of modern tools for 3D image processing,adapted for needs and requirements of automatic ultrasonic testing and code based

evaluation procedures.

Figure 20: 3D representation of inspection results


16/16

6. Conclusion

Three inspection techniques were evaluated with respect to the immersion testing of CFKcomponents and the detection of characteristic defects like delaminations, porosity andfilms.

The results obtained by Sampling Phased Array method and the single elementtransducer in contact mode are comparable and provide good results, while conventionalphased array provided poor detectability of defects due to its low sensitivity and resolution.

It is noteworthy that, the single element probe is specially designed for testing ofthin CFK-components and can be only used for limited tasks. The use of this type oftransducers for industrial applications would be difficult due to reasons of limitedinspection range, small effective aperture and consequential limited coverage of theinspection volume. Further this probe cant be used for radii inspection or testing objects ofcomplex shapes.

The Phased Array transducers with Sampling Phased array technique can replaceconventional phased array techniques. It provides reasonably good inspection results and

offers all advantages of phased array technique, e.g. for radii inspection in immersion andcontact. Furthermore Sampling Phased Array technology offers improved detectability ofcharacteristic defects and offers high potential for industrial implementation of quantitativeautomatic inspection systems.

The ongoing project is focussed on developing a practical solution for radiiinspection with Sampling Phased Array technique and automatic evaluation of inspectionresults.

References

[1] H. Wstenberg, G. Schenk, Entwicklungen und Trends bei der Anwendung von steuerbarenSchallfeldern in der ZfP mit Ultraschall, Mainz, DGZfP-Jahrestagung 2003

[2] Krning M., Hentschel D., von Bernus L., Bulavinov A., Reddy K. M.: Verfahren zur zerstrungsfreienUntersuchung eines Prfkrpers mittels Ultraschall, Deutsches Patent Nr. DE 10 2004 059 856 B42006.09.14

[3] Krning M., Bulavinov A., Reddy K. M., von Bernus L.: Verfahren zur zerstrungsfreien Untersuchungeines Prfkrpers mittels Ultraschall, Deutsches Patent Nr. DE 10 2005 051 781 A1 2007.05.03

[4] Bulavinov A.: Der getaktete Gruppenstrahler. Saarbrcken 2005 (Dissertation)[5] Jon F. Claerbout, Cecil and Ida Green Professor of Stanford University, EARTH SOUNDINGS

ANALYSIS: Processing versus Inversion, March 23, 2004[6] Bernus L. von, Bulavinov A., Dalichow M., Joneit D., Krning M., Reddy K.M.: Sampling Phased

Array: A New Technique for Signal Processing and Ultrasonic Imaging, In: Insight. Volume 48, Issue9, Pages 545-549,September 2006

[7] Bulavinov A., Joneit D., Kosov S., Krning M., Pinchuk R., Pudovikov S., Ramanan S.V., Reddy K.M.,Yastrebova O., Zhantlessov Y.: Die Getaktete Gruppenstrahlertechnik und ihre Anwendungen, In:Seminar Moderne Systemtechnik bei Prfungen mit Ultraschall des Fachausschusses Ultraschallder DGZfP, Wuppertal, 12. 13. November 2007

04. application of sampling phased array technique for ultrasonic inspection of cfrp components.pdf

Documents